Most important biological and biomedical processes are carried out not by isolated macromolecules but by complexes of interacting macromolecules. We are studying the structures of several such assemblies and the interactions among their components, using electron microscopy in combination with other approaches. The complexes currently under study are important for: intracellular proteolyis of defective and regulatory proteins; amyloidosis, with particular reference to the connection between amyloid formation and prion biology; clathrin-coated vesicles and their disassembly by the Hsc70 ATPase. (1) Role of Energy-dependent Proteases in Protein Quality Control and Cell Regulation. All cells must be capable of degrading aberrant and foreign proteins that would otherwise pollute the cell. Programmed degradation of regulatory factors is a major factor in controlling the cell cycle. Both activities are carried out by energy-dependent proteases, which consist of two parts - a peptidase and a chaperone-like ATPase. The archetypal complex of this kind is the proteasome. For several years, our studies focussed on the Clp proteases of E. coli, a model system with a similar repertoire. We showed that the peptidase ClpP consists of two apposed heptameric rings and the cognate ATPase - either ClpA or ClpX - is a single hexameric ring. ClpA/X stack axially on one or both faces of ClpP to form active complexes. We went on to show that substrate proteins bind to distal sites on the ATPase and are then translocated axially into the digestion chamber inside ClpP. In FY05, we continued working to reconcile our cryo-EM reconstruction of the ClpA hexamer (ATPgS state) at 1.1 nm resolution with the published crystal structure of the ClpA monomer (ADP state). Good agreement is obtained in the hexameric ring of D1 ATPase domains, which is the most static part of the structure. The D2 ring, which has higher ATPase activity, shows discrepancies in the interior, suggesting local mobility. We posit that this mobility is exploited in the processing of substrate proteins. (2) We studied the interaction of the 20S proteasome, its peptidase, with PA200, a 200-kDa nuclear protein that stimulates proteasomal hydrolysis of peptides. Monomers of PA200 bind to one or both ends of the 20S core. At 2.3 nm resolution, PA200 is seen to have an asymmetric dome-like structure with major and minor lobes. Its structure is likely to be an irregular folding of an alpha-helical solenoid composed of HEAT-like repeats. PA200 binding induces an opening of the axial channel through the alpha-ring of 20S. Thus PA200 activates via allosteric effects on the 20S core particle, perhaps facilitating release of digestion products or the entrance of substrates. (2) Amyloid Filament Formation by the Yeast Prions, Ure2p and Sup35p, and Other Amyloidogenic Proteins. Amyloid is fibrous aggregates of protein(s) in protease-resistant, beta-sheet-rich, non-native conformations that accumulate in disease situations, including rheumatoid arthritis. Prions (infectious proteins) are transmissible amyloids that have been implicated in neuropathies, including the spongiform encephalopathies. To investigate amyloids and the mechanisms that underlie their formation, we started studying the structures of yeast prions in 1998. We focused initially on Ure2p, a protein that normally functions as a negative regulator of nitrogen catabolism. Our previous work has shown that the N-terminal "prion domain" of Ure2p is responsible for filament formation, and the C-terminal domain which performs its regulatory function remains folded in the filamentous state but is inactivated by a steric mechanism. In our "amyloid backbone" model of the filament, the prion domains form a backbone surrounded by the C-terminal domains. In FY05, we performed experiments to test this model and to discriminate between it and an alternative model whereby filament formation is envisaged instead to be driven by interactions between N-domains and C-domains. Electron diffraction and X-ray diffraction were used to demonstrate that the backbone does indeed consist of polymerized N-domains in cross-beta conformation. NMR spectroscopy was used to confirm that in soluble Ure2p, the N-domain is highly mobile, indicative of complete disordering: this mobility disappears in the filamentous (amyloid) state. To further demonstrate that intracellular filaments are structurally indistinguishable from filaments assembled in vitro, we refined the electron tomograms of thin sections of infected yeast cells reported last year. In 2004, we formulated the "beta super-pleated structure" model for the amyloid species found in the backbone of Ure2p filaments. It envisages an array of parallel beta-sheets generated by stacking monomers, each of which adopts a planar "beta-serpentine" fold. The model accounts for all current data on Ure2p filaments. We also adapted this model to apply to filaments of the human protein, amylin, whose fibrillation is implicated in type 2 diabetes. This model has only three beta-strands per subunit serpentine as compared to eight or more in Ure2p, and explains why rat amylin does not fibrillize on account of its having amino acid substitutions in a few key positions compared to human amylin. We also started electron microscopic work on filaments of the fungal prion protein, Het-S which differs from Ure2p in that its prion domain does not have high concentrations of asparagine and is at the C-terminal end, not the N-terminal end, of the molecule. (3) Interaction of clathrin with proteins that regulate its assembly. The protein clathrin plays a key role in intracellular trafficking, via its polymerization into the coats of coated pits and vesicles. Assembly of clathrin is promoted by accessory proteins such as auxilin and AP180, and disassembly is effected by the Hsc70 ATPase. In the 1980s, we studied the molecular composition of coated vesicles and the plasticity of the assembly subunit, the clathrin triskelion. We have now returned to this system, equipped with cryo-EM technology, and are investigating the interaction of regulatory proteins - in particular, Hsc70 - with clathrin lattices. In FY05, we completed the current phase of our investigation into the binding of Hsc70 to clathrin baskets. The chimera C58J is a minimal construct capable of supporting both reactions, i.e assembly and disassembly. It consists of the C58 moiety of AP180, which facilitates clathrin assembly, fused with the J-domain of auxilin, which recruits Hsc70 to baskets. We studied the first steps in disassembly by using cryo-electron microscopy to identify the binding site of Hsc70 on clathrin-C58J baskets at pH 6: under these conditions, disassembly does not proceed further. Hsc70 interactions involve two sites: (i) its major interaction is with the sides of spars of the clathrin lattice, close to the triskelion hubs; and (ii) a site at the N-terminal hooks of the clathrin heavy chains, presumably via the J-domain of C58J. We have proposed that individual triskelions may be extricated from the clathrin lattice by the concerted action of up to six Hsc70 molecules, which intercalate between clathrin leg segments, prying them apart. Three Hsc70s remain bound to the dissociated triskelion, close to its trimerization hub.